Modern Countercurrent Chromatography - American Chemical Society

considered as Van-Deemter plots as the capacity factors of the .... 1-SF is equal to VM/VC and L.S is equal to Vc. Thus equation 10 may be written: ...
1 downloads 0 Views 1MB Size
Chapter 3 Orbital Turns per Theoretical Plate for Countercurrent Chromatography Device Comparison

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

J.-M. Menet, M.-C. Rolet-Menet, D. Thiébaut, and R. Rosset Laboratoire de Chimie Analytique, Unité de Recherche Associée au Centre National de la Recherche Scientifique 437, Ecole Supérieure de Physique et Chimie Industrielles, 10 Rue Vauquelin, 75231 Paris Cedex 05, France Counter-Current Chromatography involves many parameters including the geometrical characteristics of the devices and an essential one which is the volume of stationary phase retained inside the column. Studies of the plots of the number of theoretical plates versus the flow-rate of the mobile phase have showed they cannot be considered as Van-Deemter plots as the capacity factors of the separated solutes are not constant. The number of orbital turns per theoretical plate is introduced as a mean of comparison between different type J Coil Planet Centrifuges. It is applied to two Counter-Current Chromatographs, based on the planetary motion of a coiled Teflon tube and consequently named type J Coil Planet Centrifuges, using separations of saturated fatty acids and of phenols.The number of orbital turns per theoretical plate reveals as a useful comparison parameter as theflow-rateof the mobile phase, the rotational speed of the apparatus, the internal diameter and the length of the tubing constituting the column are involved. The retention of stationary phase inside the column and the number of theoretical plates are also included. The theory of plates was developed fifty years ago (i, 2). It consists in the cut of the column into a number offictiveareas where the equilibria are realized. These areas are called theoretical plates. This model can be used for countercurrent chromatography (3). Thus, the number of theoretical plates (N) is a measure of the column efficiency and of the spreading of the injection bands during the chromatographic process. The shape of elution peaks may be approximated by a normal distribution. To compare columns of various lengths, the height equivalent to a theoretical plate may be used. In HPLC the plot of Ν versus the mobile phaseflow-rate(4) (Van Deemter curve) using a retained solute which capacity factor is constant in the operating conditions shows a maximum at the so-called optimum flow-rate. Reports of plots of the number of theoretical plates in High Speed CounterCurrent Chromatography (HSCCC) versus the mobile phaseflow-rate(5,6) pointed out the problem of the variation of the stationary phase retention during the 0097-6156/95/0593-0035$12.00/0 © 1995 American Chemical Society In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

MODERN COUNTERCURRENT CHROMATOGRAPHY

36

experiments. As a result, the increase of the mobile phase flow-rate led to variations of the solutes capacity factors. Thus, no real comparison could be done with HPLC. The comparison of type J Coil Planet Centrifuges (CPCs) between each others is difficult because of their various geometrical properties (length, i.d. of the column, number of columns, etc. ); Ν or H do not take into account the rotational speed of the apparatus. Thus, in this paper we propose to introduce the count of orbital turns per plate (n) in order to compare CPCs and to investigate the efficiency of the process from the point of view of a time scale, η being the rotational speed times the reversal of the efficiency per unit of time.

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

Theoretical Background Due to the geometrical properties of the Gaussian distribution, the number of theoretical plates (TP) can be calculated using equation (4): Ν = 16 (t /W) = 5.54 (tf/δ) 2

2

r

(1 )

where t is the retention time of the solute used to determine N , W and δ are, respectively the peak width at the base and at one-half the height of the peak in time units. The second formula is better because errors in drawing tangents are avoided. The height equivalent to a TP is defined by the formula (4): r

HETP = UN = Η

(2)

where L is the length of the column. The count of orbital rounds per plate was calculated using some simple theoretical points which are described below. All the tubes used to make a column are supposed to be of a cylindrical shape. For a single-phase system, the linear velocity u in a tube was defined using the flow-rate: u = F/S

(3)

where F is the flow-rate of the phase and S the section of the tube. If the phase is taken as a perfect fluid (no viscosity), the velocity in the tube is homogeneous and equal to u. For a viscous fluid, the velocity distribution is parabolic (7). u is then the average velocity and the velocity on the axis of the cylinder is 2u, while near the wall of the tube it is equal to zero ; thus, u was used as an average velocity for a fluid in a cylindrical tube. For a biphasic system, the calculation of the velocities of each phase is rather complicated (8). The model used here is very simple. In countercurrent chromatography, the use of a biphasic system allows to define a mobile phase and a stationary phase. The latter was considered as motionless in the tube while the mobile phase was flowing through it. The model requires that each coil of the column contains stationary phase. Thus the effective section of the tube was reduced because of the stationary phase retained in the tube. It is convenient to use S F (9) (fractional volume of column occupied by stationary phase), defined as : S = Vs/V F

Vc = V + V S

(4)

C

M

(5)

where Vs is the volume of the stationary phase, V M the volume of the mobile phase

In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

3. MENET ET AL.

37

Orbital Turns per Theoretical Plate

and Vc the volume of the column. The effective section S ff of the tube is then calculated: e

S ff=S.(l-S ) e

(6)

F

where S is the real section of the tube. Thus the mobile phase is flowing through a section S ff. Its linear velocity can be written:

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

e

As this simplified model gives the velocity of the mobile phase, the time Τ required to obtain one plate is obtained using equation 8: T =

JL LMkM =

Ueff

N

( 8 )

F

The period T M of the motion of the column around the central axis of the apparatus is: Τ =1/ω

(9)

Μ

where ω is the orbital speed. Then the count of orbital rounds per plate (dimensionless number) is defined by: ^JL^SVfna TM

(10)

N

F

where L is the length of the column (in cm), Ν the number of theoretical plates, SF the fractional volume of column occupied by stationary phase, S the internal section of the tube (in cm ), F the flow-rate (in mL/min) and ω the orbital speed of the H S C C C device (in rpm). The given formula is appropriate for the calculationfromexperimental data. But it will be easier to understand the influence of each parameter by modifying it. The term 1-SF is equal to V M / V C and L.S is equal to Vc. Thus equation 10 may be written: 2

„ = ΧΜ£ = £ N F

Ν

ω

(11)

where to is the dead time, defined as ÎO=VM/F. The higher the η value, the lower the yield of the chromatograph from the point of view of speed of separation. Due to the fact that the column rotates 2 times around it own axis during a revolution, the maximum yield should occur for 2 plates per revolution, meaning that η = 0.5. Equation 11 could also be written: 1 = ^-1 toco n

(12)

This form enables to recognize the N/to parameter, which can be used for comparison between chromatographic techniques (10). The rotational speed intervenes in the definition of η because it is an intrinsic parameter of C C C .

In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

38

MODERN COUNTERCURRENT CHROMATOGRAPHY

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

Experimental Solvent Systems and Solutes. Two sets of experiments were carried out. The first one used a two-phase solvent system composed of heptane / methanol / acetic acid (1:1:1, v/v) to separate a test mixture of saturated fatty acids. The second two-phase solvent system was composed of chloroform / methanol / water (3:1:3, v/v) and was used to separate a mixture of phenols. The solvent mixtures were equilibrated at room temperature and the phases were separated shortly before use. All organic solvents were of HPLC grade. Methanol, acetic acid and heptane were purchased from Prolabo (Paris, France), other organic solvents from Rathburn (Chromoptic, Montpellier, France). All solvents were filtered before use and water was doubly distilled. Nitrogen (L'Air Liquide, Paris, France) supplied the nebulizer of the ELSD system. The paranitrophenol, phenol and stearic acid were purchased from Prolabo (Paris, France) and the two other fatty acids (myristic acid and palmitic acid) from Merck (Nogent-sur-Marne, France). Apparatus. The HSCCC apparatus consisted of two Gilson Model 303 pumps (ViUiers-le-Bel, France) for pumping the organic and the aqueous phases. The pumps were connected to a type J CPC. Two different type J CPCs were used. One was a P.C. Inc. system (Potomac, MD., USA), equipped with one column. The latter is made of a 1.6 mm i.d. PTFE tubing to give a total volume of 325 mL and it has a 161.6 m length. A counterweight is used to counterbalance the effect of the column filled with stationary and mobile phases. The β values ranged from 0.57 to 0. 85. The second type J unit was a Model CCC 3000 system (Pharma-tech Research Corp., Baltimore, MD., USA) which is equipped with three identical columns. Each one is prepared from 0.8 mm i.d. PTFE tubing. As they are mounted in series, the total capacity of this apparatus is 45 mL and the total length is 89.5 m. The β values ranged from 0.55 to 0.75. A stroboscope (Strob 1, AOIP, Paris, France), which measures are independent from the studied devices, allowed checking of the rotational speeds. Samples were injected into the columns via Rheodyne Model 7125 injection valves. The samples, previously dissolved in the mobile phase, were carried out into the mobile phase after filling the column with the stationary phase. The detection used for the separation of fatty acids was based on evaporative light-scattering detection (ELSD) on-line with HSCCC (11). The unit was a Sedex 45 ELSD system (Sédéré, Vitry-sur-Seine, France) manufactured for HPLC and used without modification. For the separation of phenols, a Model 2550 UV detector (Varian, Les Ulis, France) was used (72). Isopropanol was continuously added to the column effluent using a Model 8500 syringe pump (Varian, Les Ulis, France) via a Model 811 dynamic mixing chamber (Gilson, Villiers-le-Bel, France) in order to improve solute detectability (72). All data were stored either on Shimadzu CR 4A or CR 6 A integrators (Touzart et Matignon, Vitry s/Seine, France). Experimental Conditions. They are reported for the two CPCs in Table I for the separations of three saturated fatty acids, namely myristic acid (C14H28O2), palmitic acid (C16H32O2) and stearic acid (C18H36O2). A typical separation is shown in Figure 1. The chromatogram demonstrates very highflow-ratessuch as 21 mL/min are compatible with satisfactory resolutions and high efficiencies. The resolution between the stearic and palmitic acids (first two peaks) is 1.22 and between the palmitic and the myristic acids (last two peaks) is 1.54. The efficiencies are 1900 theoretical plates (TP) for the stearic acid peak, 1480 TP for the palmitic acid peak and 1480 TP for the myristic acid. The three saturated fatty acids are separated in less than 16 minutes.

In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

3.

MENET ET AL.

39

Orbital Turns per Theoretical Plate

Separation of phenol, paranitrophenol and orthonitrophenol is shown in Figure 2. Only paranitrophenol and phenol were used for the calculations because orthonitrophenol is not retained by the stationary phase. The second peak corresponds to 615 TP and the third one to 625 TP. The resolution between the two last peaks is 1.50. Table I. Experimental conditions for the separation of myristic, palmitic and stearic acids on the CCC 3000 and P.CInc. CPCs

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

Model

CCC 3000 P.C. Inc.

Speed

Sample loop volume

ω

V

Temperature

(rpm)

(mL)

1700 750

20 450

ELSD conditions Gain

CO

Nitrogen Pressure (Bars)

25 25

2 2

7 7

Results and Discussion Retention of Stationary Phase. Sp (defined in equation 4) is a measure of the volume of stationary phase retained inside the tubing of the column. The variation of Sp versus the flow-rate was studied on the P.C. Inc. with the heptane / acetic acid / methanol (1:1:1, v/v) and the chloroform / methanol / water (3:1:3, v/v) systems as shown in Figures 3A and 3B. The plots are similar in shape, as Sp linearly decreases with the flow-rate. The higher the flow-rate, the more the stationary phase is driven by the mobile phase, hence the decrease in the retention of stationary phase. However, the shape is changed for low flow-rates when using the CCC 3000 HSCCC with the heptane / acetic acid / methanol (1:1:1, v/v) solvent system. The plot, shown in Figure 3C, is a plateau below 2 mL/min before showing the same linear decrease observed on the P.C. Inc. apparatus. One explanation could be the higher running rotational speed combined with a smaller internal diameter of the coil tubing reducing the role of the drag of the stationary phase by the mobile one. Effective Linear Velocity of the Mobile Phase. Ueff is calculated from equation 7 for the two different separations achieved on the P.CInc. (Figures 4A and B) and that achieved on the CCC 3000 (Figure 4C). All the plots display the same shape, which is an increasing curve, quite horizontal at higher flow-rates. According to equation 7, Ueff tends toward the asymptotic value u (defined in equation 3) for increasing flow-rates. This means that increasing a small flow-rate significantly increases u ff, while an increase of an already high flow-rate does not change u ff so much. e

e

Efficiency. Three plots of the efficiency (number of theoretical plates) versus the flow-rate of the mobile phase are given. Two of them were achieved on a P.C. Inc. apparatus; one is based on the separation of the mixture of three fatty acids and is shown in Figure 5A. The other one involves the separation of phenols and is reported in Figure 5B. The same mixture of fatty acids was also separated on the CCC 3000 and the corresponding plot is displayed in Figure 5C. Using the same test mixture of saturated fatty acids, the higher efficiencies are obtained wilh the CCC 3000 CPC. The number of theoretical plates ranges from 500 to 2500, while it ranges from 500 to 2000 for the P.C. Inc. unit. However, the total lengths of the columns are different between these type J CPCs. For this reason, it is necessary to introduce the length of the column for a first comparison of the devices.

In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

40

MODERN COUNTERCURRENT CHROMATOGRAPHY

16 (min)

12

Figure 1. On-line H S C C C - E L S D chromatogram of a test mixture of fatty acids. P.C. Inc. apparatus; mobile phase: heptanerich;flow-rate: 21 mlVmin; concentration: 1 g/L for each fatty acid; S F = 0.51. Other experimental conditions described in Table I.

20

40

60

(min)

Figure 2. On-line H S C C C - U V chromatogram of a mixture of phenols: (1) orthonitrophenol, (2) paranitrophenol and (3) phenol. P.C. Inc. apparatus; mobile phase: choroform / methanol; flow-rate: 2.5 mL/min; concentration: paranitrophenol: 0.7 g/L and phenol: 1.6 g/L; S F = 0.86. U V detection: 2.56 A . U . , 270 nm. Isopropanol flow-rate: 0.5 mL/min. In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

3.

Orbital Turns per Theoretical Plate

MENET ET AL.

41

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

u.

0,4 H—i—i—i 0

1—ι—-—i—i—ι—ι—-—-—i—«—ι—-—«—·—ι—ι—i—-—ι 5

10

15

20

1—ι 25

Flow-rate (mL/min)

(0

(0

0,5 H 0

->

1



1

2

ι

1

ι



3

1

4

Flow-rate (mUmin)

Figure 3. Plot of the retention of stationary phase (Sp) versus the flow-rate. A) P.CInc. apparatus, heptane / acetic acid / water system (1:1:1, v/v); B) P.CInc. apparatus, chloroform / methanol / water system (3:1:3, v/v); C) CCC 3000 apparatus, heptane / acetic acid / water system (1:1:1, v/v). In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

MODERN COUNTERCURRENT CHROMATOGRAPHY

0

1

2

3

4

Flow-rate (mL/min)

Figure 4. Plot of the efficicient linear velocity of the mobile phase (ueff) versus the flow-rate. A) P.CInc. apparatus, heptane / acetic acid / water system (1:1:1, v/v); B) P.CInc. apparatus, chloroform / methanol / water system (3:1:3, v/v); C) CCC 3000 apparatus, heptane / acetic acid / water system (1:1:1, v/v). In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

43

Orbital Turns per Theoretical Plate

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

MENET ET AL.

0 H 0

1

1

1

1

2

1

1

1

1

3

1

4

Flow-rate (mL/min)

Figure 5. Number of theoretical plates (N) versus the flow-rate (F). A) P.CInc. apparatus, heptane / acetic acid / water system (1:1:1, v/v), mobile phase: heptane rich, separation of myristic, palmitic and stearic acids; B) P.CInc. apparatus, chloroform / methanol / water system (3:1:3, v/v), mobile phase: chloroform / methanolrich,separation of orthonitrophenol, paranitrophenol and phenol; C) CCC 3000 apparatus, heptane / acetic acid / water system (1:1:1, v/v), mobile phase: heptanerich,separation of myristic, palmitic and stearic acids. In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

44

MODERN COUNTERCURRENT CHROMATOGRAPHY

For a 2 mL/min flow-rate, the heights equivalent to a theoretical plate are calculated for the palmitic acid (Ci6>. The P.C. Inc. device leads to 17.2 cm (i.e., 5.8 plates per meter) and the CCC 3000 unit to 7.05 cm (le., 14.2 plates per meter). The latter, with a higher operation rotational speed and a smaller internal diameter (see Table I), is twice as efficient as the other type J CPC. When compared to the average length of one coil, the height equivalent to a theoretical plate represents 40% of the length of one coil for the P.C. Inc. unit which is similar to the 45% for the CCC 3000 CPC. The shapes of the curves have already been described by Foucault et al (5) and Menet et al (6). The efficiency Ν (number of theoretical plates) usually shows a minimum versus the flow-rate F, when the column is equilibrated at its maximum capacity for retaining the stationary phase (i.e., hydrodynamic equilibrium). However, all these curves shall not be called Van-Deemter plots. The hydrodynamic equilibrium, which is flow-rate dependent, determines the retention of stationary phase, as shown in Figure 3. Consequently, the volume of stationary phase retained in the column, hence the capacity factors, vary during the calculation of Ν vs. F, whereas Van-Deemter plots are obtained at constant capacity factors. Such results imply that the flow-rate can be increased in order to raise the efficiency and to shorten the separation time, but with a reduced retention of stationary phase and consequently a diminished resolution (5,6). An optimum has to be determined between efficiency and resolution. Count of Orbital Turns per Plate. In order to take into account the separation time and the mechanical characteristics of the apparatus, we propose to introduce n, the count of orbital turns per plate calculated according to equation 10 or 11; η can be calculated for each solute which is separated. The length of the coiled Teflon tube L and its section S are geometric constants. The flow-rate of the mobile phase F and the rotational speed of the apparatus ω are experimental parameters, chosen for each separation. But the number of theoretical plates Ν and the retention of stationary phase SF have to be calculated after the separation. The calculations of η are derived from Figures 3 and 5 showing the dependence of the retention of stationary phase and of the number of theoretical plates on the flowrate. Figures 6 A and Β represent the variation of η versus the flow-rate of the mobile phase on the P.C. Inc. unit for a separation of fatty acids and a separation of phenols. Figure 6C shows the same plot achieved on the CCC 3000 CPC for the separation of fatty acids. The shapes of the curves are similar for both type J CPCs and for two different separations achieved on the P.CInc. apparatus. It is a decreasing curve which tends to a plateau. It means that all the units used for this work are able to reach the same increased mechanical yield (below 10 turns per plate) despite their differences (column length and number, internal diameter or orbital and planet radius) at their higher flowrates. But the lower the internal diameter of the column, the lower the value of the flow-rate required to obtain the same η value: for example, at η = 10, the flow-rate is 8 -10 ml/min for the P.CInc. and 2.5 ml/min for the CCC 3000. In that case, both CPCs lead to the same u ff, approximately 28 cm/s. Using the same mixture of fatty acids at a 2 mL/minflow-rate,η is calculated for the palmitic acid. The value is 28 for the P.C. Inc. unit (ueff=8 cm/s) and 15 for the CCC 3000 device (ueff=21 cm/s). The latter is more efficient than the other CPC, but the increase in efficiency is reduced compared to the classification based on the height equivalent to a theoretical plate. The reason is that η involves the rotational speed and the internal diameter (via the section of the tube) which allows a more precise comparison of the type J CPCs. e

Conclusion Using mixtures of saturated fatty acids or phenols, separations carried out on

In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

Figure 6. Count of orbital turns per plate (n) versus the flow-rate (F). A), B) and C) defined in the caption of figure 4. In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.

MODERN COUNTERCURRENT CHROMATOGRAPHY

Downloaded by UNIV OF GUELPH LIBRARY on June 28, 2012 | http://pubs.acs.org Publication Date: May 15, 1995 | doi: 10.1021/bk-1995-0593.ch003

46

the two different type J CPCs, namely the P.CInc. and the CCC 3000, allowed to study the influence of the flow-rate on the efficiency. Plotting the number of theoretical plates versus the flow-rate revealed a reversed curve compared to classical VanDeemter curves. However, these plots are dependent on the retention of stationary phase and for Counter-Current Chromatography, they shall not be called Van-Deemter plots. A better comparison of the devices is based on the use of the height equivalent to a theoretical plate. However, we needed a parameter including the rotational speed, the internal diameter of the column as well as the retention of the stationary phase inside the column. The number of orbital turns per theoretical plate has revealed as a good comparison parameter as many typical parameters of CCC are involved. The higher flow-rates lead to the best efficiencies. Nevertheless, it should be kept in mind that high flow-rates decrease the volume of stationary phase inside the column and consequently the resolution. Acknowledgments The authors of this paper wish to thank the RHONE-POULENC RORER Company for its financial support. Literature Cited

(1) Martin, A.J.P.; Synge, R.L.M. Biochem. J. 1941,35,1358. (2) Mayer, S.N.; Tompkins, E.R. J. Amer. Chem. Soc. 1952, 52, 238. (3) Conway, W.D. Countercurrent Chromatography; Apparatus, Theory an Applications; VCH: New-York, NY, 1990; p195. (4) Rosset, R.; Caude, M.; Jardy A. Chromatographies en Phases Liquide et Supercritique; Masson: Paris, 1991. (5) Foucault, A. P.; Bousquet, Ο.; Le Goffic, F. J. Liq. Chromatogr. 1992, 15, 2691-2706. (6) Menet, J.-M.; Rolet, M.-C.; Thiébaut, D.; Rosset R.; Ito, Y. J. Liq. Chromatogr. 1992, 15, 2883-2908. (7) Poiseuille, J.L. Comptes rendus 1840, 11, 961 and 1041. (8) Govier, G.W.; Aziz, K. The Flow of Complex Mixtures in Pipes; Robert Ε. Krieger Publishing Co., Inc.: New-York, NY, 1977. (9) Ibid (3), except p198. (10) Caude, M.; Rosset, R. Analusis 1986, 14 (6), 310-311. (11) Drogue, S.; Rolet, M.-C.; Thiébaut, D.; Rosset, R. J. Chromatogr. 1991, 538, 91-97. (12) Rolet, M.-C.Thèse de l'Université Paris VI. 1993. RECEIVED November 28, 1994

In Modern Countercurrent Chromatography; Conway, W., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 1995.